Correlation at each band

Models 50 m band
Models using functional groups
Model Comparisons
Model 1: Original
## real.visitation ~ focalmonth..50 * Height + offset(lograte) +
## (1 + focalmonth..50 | Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0405307
|
0.1886986
|
0.2147905
|
0.8299306
|
|
focalmonth..50
|
0.5152201
|
0.2956979
|
1.7423871
|
0.0814407
|
|
Heightlow
|
0.7238617
|
0.0437088
|
16.5610200
|
0.0000000
|
|
focalmonth..50:Heightlow
|
-0.2752461
|
0.0374950
|
-7.3408839
|
0.0000000
|
Model 2: Trees without any visits from Dispersers are removed
## real.visitation ~ focalmonth..50 * Height + offset(lograte) +
## (1 + focalmonth..50 | Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0415496
|
0.1880433
|
0.2209575
|
0.8251255
|
|
focalmonth..50
|
0.5152245
|
0.2960341
|
1.7404230
|
0.0817848
|
|
Heightlow
|
0.7235774
|
0.0436765
|
16.5667539
|
0.0000000
|
|
focalmonth..50:Heightlow
|
-0.2750984
|
0.0374846
|
-7.3389674
|
0.0000000
|
Results are qualitatively similar.
Model 3: Original amount of trees, but fruiting neighborhood is now binary
## real.visitation ~ bin50 * Height + offset(lograte) + (1 + bin50 |
## Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0439227
|
0.1998483
|
0.219780
|
0.8260425
|
|
bin501
|
0.5713877
|
0.3322725
|
1.719636
|
0.0854987
|
|
Heightlow
|
0.6624992
|
0.0483892
|
13.691042
|
0.0000000
|
|
bin501:Heightlow
|
-0.3084512
|
0.0692597
|
-4.453547
|
0.0000084
|
Again, binarizing is qualitatively similar.
Model 4: Original amount of trees, but fruiting neighborhood categorized into 0,1,2,3+
## real.visitation ~ bin50 * Height + offset(lograte) + (1 + bin50 |
## Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0276333
|
0.1891375
|
0.1461019
|
0.8838409
|
|
bin50
|
0.5471455
|
0.2952029
|
1.8534560
|
0.0638170
|
|
Heightlow
|
0.7421483
|
0.0450024
|
16.4913175
|
0.0000000
|
|
bin50:Heightlow
|
-0.3186904
|
0.0420803
|
-7.5733801
|
0.0000000
|
Model 5: Original data but tree with outlier in visitation in high cameras removed
## real.visitation ~ focalmonth..50 * Height + offset(lograte) +
## (1 + focalmonth..50 | Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0650785
|
0.1908849
|
0.3409307
|
0.7331558
|
|
focalmonth..50
|
0.3663191
|
0.3096886
|
1.1828626
|
0.2368636
|
|
Heightlow
|
0.6933769
|
0.0439534
|
15.7752735
|
0.0000000
|
|
focalmonth..50:Heightlow
|
-0.0700859
|
0.0400875
|
-1.7483212
|
0.0804084
|
Focalmonth..50 and the interaction term are no longer significant.
Model 6: Trees without any visits from Dispersers are removed AND outlier in visitation in high cameras removed
## real.visitation ~ focalmonth..50 * Height + offset(lograte) +
## (1 + focalmonth..50 | Tree)
|
|
Estimate
|
Std. Error
|
z value
|
Pr(>|z|)
|
|
(Intercept)
|
0.0657147
|
0.1905568
|
0.3448563
|
0.7302024
|
|
focalmonth..50
|
0.3664713
|
0.3096947
|
1.1833307
|
0.2366781
|
|
Heightlow
|
0.6931354
|
0.0439257
|
15.7797227
|
0.0000000
|
|
focalmonth..50:Heightlow
|
-0.0699615
|
0.0400800
|
-1.7455466
|
0.0808898
|
Results are qualitatively similar again when trees with 0 disperser visitation is removed